Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Release v1.6.0 #600

Merged
merged 86 commits into from
Oct 4, 2022
Merged

Release v1.6.0 #600

merged 86 commits into from
Oct 4, 2022

Conversation

heyealex
Copy link
Contributor

@heyealex heyealex commented Oct 4, 2022

Submission Checklist

  • Have you installed and run this change against pre-commit? (pre-commit install)
  • Are all tests passing? (make tests)
  • Have you written unit tests to cover this change?
  • Is unit test coverage still above 80%?
  • Have you updated all applicable documentation?
  • Have you followed the guidelines in our Contributing document?

tpdownes and others added 30 commits September 12, 2022 15:22
When used to provision A2 machine types, using the empty list as the
default value of guest_accelerator causes a permadiff because the
Compute Engine API reports the A2's A100 GPU card and we have chosen an
explicit value. If, instead, we use a null value for guest_accelerator,
Terraform treats the difference as not requiring action.
Add customized version output for debugging improvement
Exposing enable_reconfigure in Slurm-onGCP V5
Adding Slurm on GCP V4 static nodes functionality
Previously the scope for the vm only allowed it to read from GCS. Now both read and writes are allowed.
Set enable_smt default to false for slurm v5 modules
Updates the schedmd-slurm-gcp-v5-hybrid module to default the
`slurm_cluster_name` to a filtered version of the `deployment_name` if
not explicitly provided. Implementation matches that of login and
controller modules.
…-disk

Add auto-delete boot disk as an option on vm-instance
Default slurm_cluster_name to deploy name in hybrid
Default scope now allows reading AND writing.
Updated the install path sed command to also catch EpilogSlurmctld and
PrologSlurmctld when exclusive partitions are being used.

Updated the hybrid test config to include an exclusive partition and to
reformat module definitions in line with recent improvements in the
codebase.
Upgrade DDN-EXAScaler to v6.1.0
heyealex and others added 23 commits September 19, 2022 23:04
Creation of the group level directory was skipped previously, as it was
handled in non-git sources automatically. With git sources, it needs to
be created, so this commit adds that logic before handling modules
directly in copySource.
Enable DDN Lustre client install with pre-existing-network-storage
Changes the deployment_name and therefore slurm_cluster_name so that the
hostname of the controller will be cleaner looking and allow us to
include the name directly in the instructions.

In addition, other reviewer feedback is addressed by this commit.
Breaks the group directory creation into a separate function from
copySource. This allow for better testing and better separation of
tasks. Tests have been written to ensure directories are created
correctly.
Add hybrid slurm demo example using 2 GCP projects.
Use fully-qualified Ansible module names where not used
Perform regular cleanup of Filestore VPC network peerings
- do not attempt to re-create a Spack environment that already exists
Fix filestore peering network cleanup script
@nick-stroud nick-stroud assigned heyealex and unassigned nick-stroud Oct 4, 2022
@heyealex heyealex merged commit 54270c1 into main Oct 4, 2022
@heyealex heyealex deleted the release-candidate branch October 4, 2022 23:27
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

7 participants